- Sport Business
- Members Log In
“Everyone is going to make decisions that are biased in their nature but sometimes you’re aware of it and hopefully you don’t fall into that trap too often.”
Karl Cooke, Performance Team Director at the Western Australian Institute of Sport, adds that groupthink is a constant threat. “It’s human nature and there’s a whole bunch of things that can cause us to make bad decisions,” he observes. “Sometimes in the heat of the battle, you don’t have time to take in all this information.”
He then poses a key question: “Can we use technology to cut out the noise and present our teams with pertinent information?” Here, we look at some of the ways data can be used to ensure your team does not lose sight of the outcome or goal.
Always challenge conventional thinking
It is hard to argue against firm evidence and, in the right hands, it can be used to challenge the sort of groupthink that can stem from a collective approach. “Every sport will have its conventions,” says Cooke. “We build metrics of say defence or attack and, to take the former, you build up intelligence on the really important things for evaluating defensive performance. You start to blend the knowledge and experience of key coaches with an evidence base from the sport.”
Ryan Murray, Director of Baseball Analytics at the Texas Rangers, says: “People are making decisions based on the information you are providing them. Say the coach wants a visual of the strike zone by batting average, my first response to that is to tell them that it’s not very predictive of anything.” He explains it to them in terms that never lose sight of the overarching goal. “Let’s actually break down into what the coach is trying to understand and how we can get to that goal. I might tell them that our internal metric, a combination of exit velocity and launch angle and hit direction, is much more predictive. You try to meet them on their plain and make it an ongoing conversation. If they ask ‘why is this section highlighted red?’ we can really open up a dialogue that allows you to do build that trust with the people who are ultimately making the decisions.”
At Harlequins, says Batchelor: “We have the sort of environment where rather than go for something new, we’re going to go for something better. It might turn out that that better thing is something that might have existed for a long time but data and information helps to put proof to it.”
Challenge yourself in what you collect… and be ready to be challenged by others
Whether it is coaches or support staff, your decision-making is going to be better informed with a workforce ready and willing to challenge the status quo. They can also help ensure the data stays focused on performance and not the hunch of a data scientist. “I can dig and I can look,” says Chad Gerhard, the former Applied Sports Scientist at the Orlando Magic, “but you can very quickly disappear down a rabbit hole and it doesn’t matter. I’d much prefer people to prod and poke me. It happens a lot here, whether it’s the medical staff or the strength & conditioning staff, and it’s good for me to be challenged.”
At Harlequins, Batchelor finds himself taking the role of devil’s advocate on a regular basis: “I end up being not anti-data but ‘anti some of the reasons that are brought for data’ because there are a lot of examples out there of bad data and you have got to be aware of false positives and self-fulfilling prophecies.” He adds that much of it can come down to recruitment. “From the off you deliberately try to not just recruit people who completely agree with your perspective. We certainly have a lot of conversations in our team from different perspectives on subjects without individuals pulling it in different directions.”
It is an approach that has also served David Martin well during his time working in the NBA with the Philadelphia 76ers. “Those group meetings can be messy, emotional and sometimes argumentative but I welcome the chaos and like to brainstorm across a lot of different areas.” Ultimately, Martin’s team will disagree and commit. “There will be a delivery to coaching staff and players and those messages need to be simplified and they have to have uniform support.”
Data models can also be created that challenge convention
It is not only staff but analytic predictive models that can serve to challenge conventional thinking. “Machine learning has made things more interesting for those using predictive models,” observes Cooke. “A traditional model would have been our experience telling us that an athlete is vulnerable to injury when these things happen. Now, with tools like machine learning, over time we can train the algorithms to say these are the red flags for this athlete; and they’re evidence-based from a database of training and injury inflammation. That’s exciting in that we have the traditional model there and then sitting alongside that maybe a more objective machine learning-based to challenge our traditional thinking. We might continue to stick with the traditional model but at least we’re challenging our thinking with some other ideas as well.”
On occasion, it can lead to a revised approach, as was the case at the Texas Rangers, where Murray introduced the team’s first defensive algorithm prior to the 2015 offseason and the result was an increased number of outs that pushed the team to the division championship. “We taught coaches where the defenders need to stand at any given point on the field,” he recalls. “This was something that had never been automated before; the coaches talked about defensive positions and I wasn’t trying to replace that functionality – I was trying to augment it. We were doing analysis throughout 2015 and we decided that the algorithm was bringing us roughly 182 outs over the course of a season, which is 1.1 outs per game, all from being in the correct positions.”
This exclusive feature has been extracted from our latest Special Report: Navigating the Data Maze. Download the full report by clicking below, and keep an eye out for our next Special Report landing in just a few weeks time.
The Sport Performance Summit London
Data & Innovation is one of the four key content pillars that will be covered at our flagship event this November 12-13 to prepare our members to out-think, out-prepare and out-perform the competition. Click here to find out more about the event.